As a conventional coding method of digital time-varying image signals, there can be cited the coding method in ITU-T (International Telecommunication Union-Telecommunication Standardization Sector) Recommendation H.261 that was recommended in March, 1993. The featured coding method of H. 261 is a motion compensation prediction scheme. The motion compensation prediction scheme is performed as follows.
First, an input image picture and a pre-coded picture are compared, and a motion quantity between them is measured (motion detection). An input image picture is predicted on the basis of the motion quantity and the pre-coded picture. The difference between the predicted image (prediction image) and the input image picture (prediction error signal) is determined, and the prediction error signals and the motion quantity are transmitted to a reception side. Thereby, image information can be transmitted in a small data quantity.
Moreover, in H.261, there is another method different from the motion compensation picture prediction scheme. The method does not use the difference between the prediction image and the input image picture, but the method codes the input image picture itself. The method is called “intra-coding scheme”. The intra-coding scheme is used when no pre-coded pictures exist such as the beginning of communication although the generated data quantity is larger than that in the motion compensation prediction scheme, and where the correlation between an input image picture and a pre-coded picture is low and the data quantity is larger than that in the motion compensation prediction scheme.
Moreover, when an transmission error occurs on the transmission lines and a section of images is deteriorated, because the deterioration of the picture quality is propagated to the next picture in turn in the motion compensation prediction scheme in which only the prediction error signal is transmitted, the deterioration of the picture quality can be recovered by using the intra-coding.
Both the motion compensation prediction scheme and the intra-coding scheme divide a picture into blocks (coded blocks), each having 16×16 pixels, and are applied on a per coded block basis. A picture in which all of the coded blocks in it are forcedly intra-coded is called an “intra-picture”, and a picture coded in conformity with the motion compensation prediction scheme is called an “inter-picture”.
On the other hand, there is a concealment technique for suppressing the deterioration of the picture quality when a transmission error occurs during transmission. The concealment technique is the processing to do voluntarily on the side of the reception, and is not included in the contents of the recommendation. However, by performing the processing, the picture quality deterioration in case of the occurrence of the transmission error can be suppressed. As one of the concealment techniques, there is a method in which when a transmission error occurred and received coded block data can not be decoded correctly, coded block data existing at the same position in the pre-coded picture is output as it is.
Moreover, as one of the concealment techniques, there is a method in which a motion quantity of a coded block that could correctly be decoded and exists at a peripheral position is adopted as the motion quality of the present coded block and a prediction image is generated from a pre-coded picture by the use of the adopted motion quantity and the generated prediction image is output.
When image communication begins, generally, an intra-picture is used as the first picture. This is because a pre-coded picture that becomes always necessary in case of motion compensation prediction coding does not exist at this time.
When a transmission error occurs during the transmission of the first picture, a coding block of a section of the first picture cannot be decoded correctly. In this case, because no pre-coded picture exists, it is impossible to do the concealment processing using a section of the pre-coded picture. Accordingly, the concealment processing of the section of the first picture must be performed by a prescribed value.
As described above, when a transmission error occurs in the first picture, the picture quality deteriorates more severely than the picture quality when concealment processing is performed when a pre-coded picture exists. Moreover, because the motion compensation prediction scheme is used, there is a problem that the deterioration of the picture quality in the first picture is propagated in the second and following pictures with the elapse of time.